Prediction of lower limb joint angles and moments during gait using artificial neural networks

The future of in-field sports biomechanics: wearables plus modelling compute real-time in vivo tissue loading to prevent and repair musculoskeletal injuries

This paper explores the use of biomechanics in identifying the mechanistic causes of musculoskeletal tissue injury and degeneration. It appraises how biomechanics has been used to develop training programmes aiming to maintain or recover tissue health. Tissue health depends on the functional mechanical environment experienced by tissues during daily and rehabilitation activities. These environments are the result of the interactions between tissue motion, loading, biology, and morphology. Maintaining health of and/or repairing musculoskeletal tissues requires targeting the "ideal" in vivo tissue mechanics (i.e., loading and deformation), which may be enabled by appropriate real-time biofeedback. Recent research shows that biofeedback technologies may increase their quality and effectiveness by integrating a personalised neuromusculoskeletal modelling driven by real-time motion capture and medical imaging. Model personalisation is crucial in obtaining physically and physiologically valid predictions of tissue biomechanics. Model real-time execution is crucial and achieved by code optimisation and artificial intelligence methods. Furthermore, recent work has also shown that laboratory-based motion capture biomechanical measurements and modelling can be performed outside the laboratory with wearable sensors and artificial intelligence. The next stage is to combine these technologies into well-designed easy to use products to guide training to maintain or recover tissue health in the real-world.

Exploring the contribution of joint angles and sEMG signals on joint torque prediction accuracy using LSTM-based deep learning techniques

Machine learning (ML) has been used to predict lower extremity joint torques from joint angles and surface electromyography (sEMG) signals. This study trained three bidirectional Long Short-Term Memory (LSTM) models, which utilize joint angle, sEMG, and combined modalities as inputs, using a publicly accessible dataset to estimate joint torques during normal walking and assessed the performance of models, that used specific inputs independently plus the accuracy of the joint-specific torque prediction. The performance of each model was evaluated using normalized root mean square error (nRMSE) and Pearson correlation coefficient (PCC). Each model's median scores for the PCC and nRMSE values were highly convergent and the bulk of the mean nRMSE values of all joints were less than 10%. The ankle joint torque was the most successfully predicted output, having a mean nRMSE of less than 9% for all models. The knee joint torque prediction has reached the highest accuracy with a mean nRMSE of 11% and the hip joint torque prediction of 10%. The PCC values of each model were significantly high and remarkably comparable for the ankle (∼ 0.98), knee (∼ 0.92), and hip (∼ 0.95) joints. The model obtained significantly close accuracy with single and combined input modalities, indicating that one of either input may be sufficient for predicting the torque of a particular joint, obviating the need for the other in certain contexts.

Predicting Knee Joint Contact Forces During Normal Walking Using Kinematic Inputs With a Long-Short Term Neural Network

Knee joint contact forces are commonly estimated via surrogate measures (i.e., external knee adduction moments or musculoskeletal modeling). Despite its capabilities, modeling is not optimal for clinicians or persons with limited experience. The purpose of this study was to design a novel prediction method for knee joint contact forces that is simplistic in terms of required inputs. This study included marker trajectories and instrumented knee forces during normal walking from the "Grand Challenge" (n = 6) and "CAMS" (n = 2) datasets. Inverse kinematics were used to derive stance phase hip (sagittal, frontal, transverse), knee (sagittal, frontal), ankle (sagittal), and trunk (frontal) kinematics. A long-short term memory network (LSTM) was created using matlab to predict medial and lateral knee force waveforms using combinations of the kinematics. The Grand Challenge and CAMS datasets trained and tested the network, respectively. Musculoskeletal modeling forces were derived using static optimization and joint reaction tools in OpenSim. Waveform accuracy was determined as the proportion of variance and root-mean-square error between network predictions and in vivo data. The LSTM network was highly accurate for medial forces (R2 = 0.77, RMSE = 0.27 BW) and required only frontal hip and knee and sagittal hip and ankle kinematics. Modeled medial force predictions were excellent (R2 = 0.77, RMSE = 0.33 BW). Lateral force predictions were poor for both methods (LSTM R2 = 0.18, RMSE = 0.08 BW; modeling R2 = 0.21, RMSE = 0.54 BW). The designed LSTM network outperformed most reports of musculoskeletal modeling, including those reached in this study, revealing knee joint forces can accurately be predicted by using only kinematic input variables.

Predicting ankle and knee sagittal kinematics and kinetics using an ankle-mounted inertial sensor

The purpose of this study was to develop a machine learning model to reconstruct time series kinematic and kinetic profiles of the ankle and knee joint across six different tasks using an ankle-mounted IMU. Four male collegiate basketball players performed repeated tasks, including walking, jogging, running, sidestep cutting, max-height jumping, and stop-jumping, resulting in a total of 102 movements. Ankle and knee flexion-extension angles and moments were estimated using motion capture and inverse dynamics and considered 'actual data' for the purpose of model fitting. Synchronous acceleration and angular velocity data were collected from right ankle-mounted IMUs. A time-series feature extraction model was used to determine a set of features used as input to a random forest regression model to predict the ankle and knee kinematics and kinetics. Five-fold cross-validation was performed to verify the model accuracy, and statistical parametric mapping was used to determine the difference between the predicted and experimental time series. The random forest regression model predicted the time-series profiles of the ankle and knee flexion-extension angles and moments with high accuracy (Kinematics: R2 ranged from 0.782 to 0.962, RMSE ranged from 2.19° to 11.58°; Kinetics: R2 ranged from 0.711 to 0.966, RMSE ranged from 0.10 Nm/kg to 0.41 Nm/kg). There were differences between predicted and actual time series for the knee flexion-extension moment during stop-jumping and walking. An appropriately trained feature-based regression model can predict time series knee and ankle joint angles and moments across a wide range of tasks using a single ankle-mounted IMU.

Multibody Model with Foot-Deformation Approach for Estimating Ground Reaction Forces and Moments and Joint Torques during Level Walking through Optical Motion Capture without Optimization Techniques

The biomechanical-model-based approach with a contact model offers advantages in estimating ground reaction forces (GRFs) and ground reaction moments (GRMs), as it does not rely on the need for training data and gait assumptions. However, this approach faces the challenge of long computational times due to the inclusion of optimization processes. To address this challenge, the present study developed a new optical motion capture (OMC)-based method to estimate GRFs, GRMs, and joint torques without prolonged computational times. The proposed approach performs the estimation process by distributing external forces, as determined by a multibody model, between the left and right feet based on foot deformations, thereby predicting the GRFs and GRMs without relying on optimization techniques. In this study, prediction accuracies during level walking were confirmed by comparing a general analysis using a force plate with the estimation results. The comparison revealed excellent or strong correlations between the prediction and the measurements for all GRFs, GRMs, and lower-limb-joint torques. The proposed method, which provides practical estimation with low computational cost, facilitates efficient biomechanical analysis and rapid feedback of analysis results, contributing to its increased applicability in clinical settings.

Estimation of Kinetics Using IMUs to Monitor and Aid in Clinical Decision-Making during ACL Rehabilitation: A Systematic Review

After an ACL injury, rehabilitation consists of multiple phases, and progress between these phases is guided by subjective visual assessments of activities such as running, hopping, jump landing, etc. Estimation of objective kinetic measures like knee joint moments and GRF during assessment can help physiotherapists gain insights on knee loading and tailor rehabilitation protocols. Conventional methods deployed to estimate kinetics require complex, expensive systems and are limited to laboratory settings. Alternatively, multiple algorithms have been proposed in the literature to estimate kinetics from kinematics measured using only IMUs. However, the knowledge about their accuracy and generalizability for patient populations is still limited. Therefore, this article aims to identify the available algorithms for the estimation of kinetic parameters using kinematics measured only from IMUs and to evaluate their applicability in ACL rehabilitation through a comprehensive systematic review. The papers identified through the search were categorized based on the modelling techniques and kinetic parameters of interest, and subsequently compared based on the accuracies achieved and applicability for ACL patients during rehabilitation. IMUs have exhibited potential in estimating kinetic parameters with good accuracy, particularly for sagittal movements in healthy cohorts. However, several shortcomings were identified and future directions for improvement have been proposed, including extension of proposed algorithms to accommodate multiplanar movements and validation of the proposed techniques in diverse patient populations and in particular the ACL population.

3D gait analysis in children using wearable sensors: feasibility of predicting joint kinematics and kinetics with personalized machine learning models and inertial measurement units

Introduction: Children's walking patterns evolve with age, exhibiting less repetitiveness at a young age and more variability than adults. Three-dimensional gait analysis (3DGA) is crucial for understanding and treating lower limb movement disorders in children, traditionally performed using Optical Motion Capture (OMC). Inertial Measurement Units (IMUs) offer a cost-effective alternative to OMC, although challenges like drift errors persist. Machine learning (ML) models can mitigate these issues in adults, prompting an investigation into their applicability to a heterogeneous pediatric population. This study aimed at 1) quantifying personalized and generalized ML models' performance for predicting gait time series in typically developed (TD) children using IMUs data, 2) Comparing random forest (RF) and convolutional neural networks (CNN) models' performance, 3) Finding the optimal number of IMUs required for accurate predictions. Methodology: Seventeen TD children, aged 6 to 15, participated in data collection involving OMC, force plates, and IMU sensors. Joint kinematics and kinetics (targets) were computed from OMC and force plates' data using OpenSim. Tsfresh, a Python package, extracted features from raw IMU data. Each target's ten most important features were input in the development of personalized and generalized RF and CNN models. This procedure was initially conducted with 7 IMUs placed on all lower limb segments and then performed using only two IMUs on the feet. Results: Findings suggested that the RF and CNN models demonstrated comparable performance. RF predicted joint kinematics with a 9.5% and 19.9% NRMSE for personalized and generalized models, respectively, and joint kinetics with an NRMSE of 10.7% for personalized and 15.2% for generalized models in TD children. Personalized models provided accurate estimations from IMU data in children, while generalized models lacked accuracy due to the limited dataset. Furthermore, reducing the number of IMUs from 7 to 2 did not affect the results, and the performance remained consistent. Discussion: This study proposed a promising personalized approach for gait time series prediction in children, involving an RF model and two IMUs on the feet.

Integrating an LSTM framework for predicting ankle joint biomechanics during gait using inertial sensors

The ankle joint plays a crucial role in gait, facilitating the articulation of the lower limb, maintaining foot-ground contact, balancing the body, and transmitting the center of gravity. This study aimed to implement long short-term memory (LSTM) networks for predicting ankle joint angles, torques, and contact forces using inertial measurement unit (IMU) sensors. Twenty-five healthy participants were recruited. Two IMU sensors were attached to the foot dorsum and the vertical axis of the distal anteromedial tibia in the right lower limb to record acceleration and angular velocity during running. We proposed a LSTM-MLP (multilayer perceptron) model for training time-series data from IMU sensors and predicting ankle joint biomechanics. The model underwent validation and testing using a custom nested k-fold cross-validation process. The average values of the coefficient of determination (R2), mean absolute error (MAE), and mean squared error (MSE) for ankle dorsiflexion joint and moment, subtalar inversion joint and moment, and ankle joint contact forces were 0.89 ± 0.04, 0.75 ± 1.04, and 2.96 ± 4.96 for walking, and 0.87 ± 0.07, 0.88 ± 1.26, and 4.1 ± 7.17 for running, respectively. This study demonstrates that IMU sensors, combined with LSTM neural networks, are invaluable tools for evaluating ankle joint biomechanics in lower limb pathological diagnosis and rehabilitation, offering a cost-effective and versatile alternative to traditional experimental settings.

Machine learning-based prediction of hip joint moment in healthy subjects, patients and post-operative subjects

The application of machine learning in the field of motion capture research is growing rapidly. The purpose of the study is to implement a long-short term memory (LSTM) model able to predict sagittal plane hip joint moment (HJM) across three distinct cohorts (healthy controls, patients and post-operative patients) starting from 3D motion capture and force data. Statistical parametric mapping with paired samples t-test was performed to compare machine learning and inverse dynamics HJM predicted values, with the latter used as gold standard. The results demonstrated favorable model performance on each of the three cohorts, showcasing its ability to successfully generalize predictions across diverse cohorts.

Prediction of knee biomechanics with different tibial component malrotations after total knee arthroplasty: conventional machine learning vs. deep learning

The precise alignment of tibiofemoral components in total knee arthroplasty is a crucial factor in enhancing the longevity and functionality of the knee. However, it is a substantial challenge to quickly predict the biomechanical response to malrotation of tibiofemoral components after total knee arthroplasty using musculoskeletal multibody dynamics models. The objective of the present study was to conduct a comparative analysis between a deep learning method and four conventional machine learning methods for predicting knee biomechanics with different tibial component malrotation during a walking gait after total knee arthroplasty. First, the knee contact forces and kinematics with different tibial component malrotation in the range of ±5° in the three directions of anterior/posterior slope, internal/external rotation, and varus/valgus rotation during a walking gait after total knee arthroplasty were calculated based on the developed musculoskeletal multibody dynamics model. Subsequently, deep learning and four conventional machine learning methods were developed using the above 343 sets of biomechanical data as the dataset. Finally, the results predicted by the deep learning method were compared to the results predicted by four conventional machine learning methods. The findings indicated that the deep learning method was more accurate than four conventional machine learning methods in predicting knee contact forces and kinematics with different tibial component malrotation during a walking gait after total knee arthroplasty. The deep learning method developed in this study enabled quickly determine the biomechanical response with different tibial component malrotation during a walking gait after total knee arthroplasty. The proposed method offered surgeons and surgical robots the ability to establish a calibration safety zone, which was essential for achieving precise alignment in both preoperative surgical planning and intraoperative robotic-assisted surgical navigation.

Estimation of lower limb joint moments based on the inverse dynamics approach: a comparison of machine learning algorithms for rapid estimation

The aim of this study is to estimate the joint moments of the ankle, knee, and hip joints during walking. A sit-to-stand (STS) movement analysis was first performed on 20 participants with different anthropometric characteristics. Then, analysis of the dynamics of the STS motion was used to develop a biomechanical model. Decision tree (DT), linear regression (LR), support vector machine (SVM), random forest (RF), and three deep learning (DL) algorithms and deep neural network (DNN), long-short-term memory (LSTM), and convolutional neural network (CNN) are examined in this work to estimate three joint moments: ankle, knee, and hip. The results of the seven algorithms were evaluated using four statistical benchmarks: MSR, RMSE, correlation coefficient (R), and MAE to find the most accurate one. The results show that the most successful algorithms were LSTM in estimating knee, hip, and ankle joint moments using 19 and 7 inputs. The R value was 0.9990 using 19 inputs and 0.9972 using 7 inputs. The other algorithms have a correlation coefficient (R) success of 0.9902, 0.9770, 0.9884, 0.9577, 0.9786, and 0.9022 for RF, CNN, DT, DNN, SVM, and LR, respectively. The prediction of joint moments plays a crucial role in the design of the biomechanical system with the desired mechanical properties. Especially, the need has arisen to predict joint moments in a shorter time to utilize in real-time active prosthesis/orthosis controllers.

Estimation of Lower Limb Joint Angles and Joint Moments during Different Locomotive Activities Using the Inertial Measurement Units and a Hybrid Deep Learning Model

Using inertial measurement units (IMUs) to estimate lower limb joint kinematics and kinetics can provide valuable information for disease diagnosis and rehabilitation assessment. To estimate gait parameters using IMUs, model-based filtering approaches have been proposed, such as the Kalman filter and complementary filter. However, these methods require special calibration and alignment of IMUs. The development of deep learning algorithms has facilitated the application of IMUs in biomechanics as it does not require particular calibration and alignment procedures of IMUs in use. To estimate hip/knee/ankle joint angles and moments in the sagittal plane, a subject-independent temporal convolutional neural network-bidirectional long short-term memory network (TCN-BiLSTM) model was proposed using three IMUs. A public benchmark dataset containing the most representative locomotive activities in daily life was used to train and evaluate the TCN-BiLSTM model. The mean Pearson correlation coefficient of joint angles and moments estimated by the proposed model reached 0.92 and 0.87, respectively. This indicates that the TCN-BiLSTM model can effectively estimate joint angles and moments in multiple scenarios, demonstrating its potential for application in clinical and daily life scenarios.

Smooth and accurate predictions of joint contact force time-series in gait using over parameterised deep neural networks

Alterations in joint contact forces (JCFs) are thought to be important mechanisms for the onset and progression of many musculoskeletal and orthopaedic pain disorders. Computational approaches to JCFs assessment represent the only non-invasive means of estimating in-vivo forces; but this cannot be undertaken in free-living environments. Here, we used deep neural networks to train models to predict JCFs, using only joint angles as predictors. Our neural network models were generally able to predict JCFs with errors within published minimal detectable change values. The errors ranged from the lowest value of 0.03 bodyweight (BW) (ankle medial-lateral JCF in walking) to a maximum of 0.65BW (knee VT JCF in running). Interestingly, we also found that over parametrised neural networks by training on longer epochs (>100) resulted in better and smoother waveform predictions. Our methods for predicting JCFs using only joint kinematics hold a lot of promise in allowing clinicians and coaches to continuously monitor tissue loading in free-living environments.

BioMAT: An Open-Source Biomechanics Multi-Activity Transformer for Joint Kinematic Predictions Using Wearable Sensors

Through wearable sensors and deep learning techniques, biomechanical analysis can reach beyond the lab for clinical and sporting applications. Transformers, a class of recent deep learning models, have become widely used in state-of-the-art artificial intelligence research due to their superior performance in various natural language processing and computer vision tasks. The performance of transformer models has not yet been investigated in biomechanics applications. In this study, we introduce a Biomechanical Multi-activity Transformer-based model, BioMAT, for the estimation of joint kinematics from streaming signals of multiple inertia measurement units (IMUs) using a publicly available dataset. This dataset includes IMU signals and the corresponding sagittal plane kinematics of the hip, knee, and ankle joints during multiple activities of daily living. We evaluated the model's performance and generalizability and compared it against a convolutional neural network long short-term model, a bidirectional long short-term model, and multi-linear regression across different ambulation tasks including level ground walking (LW), ramp ascent (RA), ramp descent (RD), stair ascent (SA), and stair descent (SD). To investigate the effect of different activity datasets on prediction accuracy, we compared the performance of a universal model trained on all activities against task-specific models trained on individual tasks. When the models were tested on three unseen subjects' data, BioMAT outperformed the benchmark models with an average root mean square error (RMSE) of 5.5 ± 0.5°, and normalized RMSE of 6.8 ± 0.3° across all three joints and all activities. A unified BioMAT model demonstrated superior performance compared to individual task-specific models across four of five activities. The RMSE values from the universal model for LW, RA, RD, SA, and SD activities were 5.0 ± 1.5°, 6.2 ± 1.1°, 5.8 ± 1.1°, 5.3 ± 1.6°, and 5.2 ± 0.7° while these values for task-specific models were, 5.3 ± 2.1°, 6.7 ± 2.0°, 6.9 ± 2.2°, 4.9 ± 1.4°, and 5.6 ± 1.3°, respectively. Overall, BioMAT accurately estimated joint kinematics relative to previous machine learning algorithms across different activities directly from the sequence of IMUs signals instead of time-normalized gait cycle data.

Fusion of video and inertial sensing data via dynamic optimization of a biomechanical model

Inertial sensing and computer vision are promising alternatives to traditional optical motion tracking, but until now these data sources have been explored either in isolation or fused via unconstrained optimization, which may not take full advantage of their complementary strengths. By adding physiological plausibility and dynamical robustness to a proposed solution, biomechanical modeling may enable better fusion than unconstrained optimization. To test this hypothesis, we fused video and inertial sensing data via dynamic optimization with a nine degree-of-freedom model and investigated when this approach outperforms video-only, inertial-sensing-only, and unconstrained-fusion methods. We used both experimental and synthetic data that mimicked different ranges of video and inertial measurement unit (IMU) data noise. Fusion with a dynamically constrained model significantly improved estimation of lower-extremity kinematics over the video-only approach and estimation of joint centers over the IMU-only approach. It consistently outperformed single-modality approaches across different noise profiles. When the quality of video data was high and that of inertial data was low, dynamically constrained fusion improved estimation of joint kinematics and joint centers over unconstrained fusion, while unconstrained fusion was advantageous in the opposite scenario. These findings indicate that complementary modalities and techniques can improve motion tracking by clinically meaningful margins and that data quality and computational complexity must be considered when selecting the most appropriate method for a particular application.

Machine Learning for Optical Motion Capture-Driven Musculoskeletal Modelling from Inertial Motion Capture Data

Marker-based Optical Motion Capture (OMC) systems and associated musculoskeletal (MSK) modelling predictions offer non-invasively obtainable insights into muscle and joint loading at an in vivo level, aiding clinical decision-making. However, an OMC system is lab-based, expensive, and requires a line of sight. Inertial Motion Capture (IMC) techniques are widely-used alternatives, which are portable, user-friendly, and relatively low-cost, although with lesser accuracy. Irrespective of the choice of motion capture technique, one typically uses an MSK model to obtain the kinematic and kinetic outputs, which is a computationally expensive tool increasingly well approximated by machine learning (ML) methods. Here, an ML approach is presented that maps experimentally recorded IMC input data to the human upper-extremity MSK model outputs computed from ('gold standard') OMC input data. Essentially, this proof-of-concept study aims to predict higher-quality MSK outputs from the much easier-to-obtain IMC data. We use OMC and IMC data simultaneously collected for the same subjects to train different ML architectures that predict OMC-driven MSK outputs from IMC measurements. In particular, we employed various neural network (NN) architectures, such as Feed-Forward Neural Networks (FFNNs) and Recurrent Neural Networks (RNNs) (vanilla, Long Short-Term Memory, and Gated Recurrent Unit) and a comprehensive search for the best-fit model in the hyperparameters space in both subject-exposed (SE) as well as subject-naive (SN) settings. We observed a comparable performance for both FFNN and RNN models, which have a high degree of agreement (ravg,SE,FFNN=0.90±0.19, ravg,SE,RNN=0.89±0.17, ravg,SN,FFNN=0.84±0.23, and ravg,SN,RNN=0.78±0.23) with the desired OMC-driven MSK estimates for held-out test data. The findings demonstrate that mapping IMC inputs to OMC-driven MSK outputs using ML models could be instrumental in transitioning MSK modelling from 'lab to field'.

Maintaining soldier musculoskeletal health using personalised digital humans, wearables and/or computer vision

OBJECTIVES

The physical demands of military service place soldiers at risk of musculoskeletal injuries and are major concerns for military capability. This paper outlines the development new training technologies to prevent and manage these injuries.

DESIGN

Narrative review.

METHODS

Technologies suitable for integration into next-generation training devices were examined. We considered the capability of technologies to target tissue level mechanics, provide appropriate real-time feedback, and their useability in-the-field.

RESULTS

Musculoskeletal tissues' health depends on their functional mechanical environment experienced in military activities, training and rehabilitation. These environments result from the interactions between tissue motion, loading, biology, and morphology. Maintaining health of and/or repairing joint tissues requires targeting the "ideal" in vivo tissue mechanics (i.e., loading and strain), which may be enabled by real-time biofeedback. Recent research has shown that these biofeedback technologies are possible by integrating a patient's personalised digital twin and wireless wearable devices. Personalised digital twins are personalised neuromusculoskeletal rigid body and finite element models that work in real-time by code optimisation and artificial intelligence. Model personalisation is crucial in obtaining physically and physiologically valid predictions.

CONCLUSIONS

Recent work has shown that laboratory-quality biomechanical measurements and modelling can be performed outside the laboratory with a small number of wearable sensors or computer vision methods. The next stage is to combine these technologies into well-designed easy to use products.

A comparison of machine learning models' accuracy in predicting lower-limb joints' kinematics, kinetics, and muscle forces from wearable sensors

A combination of wearable sensors' data and Machine Learning (ML) techniques has been used in many studies to predict specific joint angles and moments. The aim of this study was to compare the performance of four different non-linear regression ML models to estimate lower-limb joints' kinematics, kinetics, and muscle forces using Inertial Measurement Units (IMUs) and electromyographys' (EMGs) data. Seventeen healthy volunteers (9F, 28 ± 5 years) were asked to walk over-ground for a minimum of 16 trials. For each trial, marker trajectories and three force-plates data were recorded to calculate pelvis, hip, knee, and ankle kinematics and kinetics, and muscle forces (the targets), as well as 7 IMUs and 16 EMGs. The features from sensors' data were extracted using the Tsfresh python package and fed into 4 ML models; Convolutional Neural Networks (CNN), Random Forest (RF), Support Vector Machine, and Multivariate Adaptive Regression Spline for targets' prediction. The RF and CNN models outperformed the other ML models by providing lower prediction errors in all intended targets with a lower computational cost. This study suggested that a combination of wearable sensors' data with an RF or a CNN model is a promising tool to overcome the limitations of traditional optical motion capture for 3D gait analysis.

Lower body kinematics estimation during walking using an accelerometer

Measuring and predicting accurate joint angles are important to developing analytical tools to gauge users' progress. Such measurement is usually performed in laboratory settings, which is difficult and expensive. So, the aim of this study was continuous estimation of lower limb joint angles during walking using an accelerometer and random forest (RF). Thus, 73 subjects (26 women and 47 men) voluntarily participated in this study. The subjects walked at the slow, moderate, and fast speeds on a walkway, which was covered with 10 Vicon camera. Acceleration was used as input for a RF to estimate ankle, knee, and hip angles (in transverse, frontal, and sagittal planes). Pearson correlation coefficient (r) and Mean Square Error (MSE) were computed between the experimental and estimated data. Paired statistical parametric mapping (SPM) t-test was used to compare the experimental and estimated data throughout gait cycle. The results of this study showed that the MSE of joint angles between the experimental and estimated data ranged from 0.04 to 24.29 and r > 0.91. Moreover, the findings of SPM indicated that there was no significant difference between the experimental and estimated data of ankle, knee, and hip angles in all three planes throughout gait cycle. The results of our research developed a more accessible, portable procedure to quantifying lower limb joint angles by an accelerometer and RF. So, such wearable-based joint angles have the potential to be used in outside-laboratory settings to measure walking kinematics.

Machine-Learning-Based Methodology for Estimation of Shoulder Load in Wheelchair-Related Activities Using Wearables

There is a high prevalence of shoulder problems in manual wheelchair users (MWUs) with a spinal cord injury. How shoulder load relates to shoulder problems remains unclear. This study aimed to develop a machine-learning-based methodology to estimate the shoulder load in wheelchair-related activities of daily living using wearable sensors. Ten able-bodied participants equipped with five inertial measurement units (IMU) on their thorax, right arm, and wheelchair performed activities exemplary of daily life of MWUs. Electromyography (EMG) was recorded from the long head of the biceps and medial part of the deltoid. A neural network was trained to predict the shoulder load based on IMU and EMG data. Different cross-validation strategies, sensor setups, and model architectures were examined. The predicted shoulder load was compared to the shoulder load determined with musculoskeletal modeling. A subject-specific biLSTM model trained on a sparse sensor setup yielded the most promising results (mean correlation coefficient = 0.74 ± 0.14, relative root-mean-squared error = 8.93% ± 2.49%). The shoulder-load profiles had a mean similarity of 0.84 ± 0.10 over all activities. This study demonstrates the feasibility of using wearable sensors and neural networks to estimate the shoulder load in wheelchair-related activities of daily living.

Developing a method for quantifying hip joint angles and moments during walking using neural networks and wearables

Quantifying hip angles/moments during gait is critical for improving hip pathology diagnostic and treatment methods. Recent work has validated approaches combining wearables with artificial neural networks (ANNs) for cheaper, portable hip joint angle/moment computation. This study developed a Wearable-ANN approach for calculating hip joint angles/moments during walking in the sagittal/frontal planes with data from 17 healthy subjects, leveraging one shin-mounted inertial measurement unit (IMU) and a force-measuring insole for data capture. Compared to the benchmark approach, a two hidden layer ANN (n = 5 nodes per layer) achieved an average rRMSE = 15% and R²=0.85 across outputs, subjects and training rounds.

An algorithm to reduce human-robot interface compliance errors in posture estimation in wearable robots

Assistive forces transmitted from wearable robots to the robot's users are often defined by controllers that rely on the accurate estimation of the human posture. The compliant nature of the human-robot interface can negatively affect the robot's ability to estimate the posture. In this article, we present a novel algorithm that uses machine learning to correct these errors in posture estimation. For that, we recorded motion capture data and robot performance data from a group of participants (n = 8; 4 females) who walked on a treadmill while wearing a wearable robot, the Myosuit. Participants walked on level ground at various gait speeds and levels of support from the Myosuit. We used optical motion capture data to measure the relative displacement between the person and the Myosuit. We then combined this data with data derived from the robot to train a model, using a grading boosting algorithm (XGBoost), that corrected for the mechanical compliance errors in posture estimation. For the Myosuit controller, we were particularly interested in the angle of the thigh segment. Using our algorithm, the estimated thigh segment's angle RMS error was reduced from 6.3° (2.3°) to 2.5° (1.0°), mean (standard deviation). The average maximum error was reduced from 13.1° (4.9°) to 5.9° (2.1°). These improvements in posture estimation were observed for all of the considered assistance force levels and walking speeds. This suggests that ML-based algorithms provide a promising opportunity to be used in combination with wearable-robot sensors for an accurate user posture estimation.

Cross-Leg Prediction of Running Kinematics across Various Running Conditions and Drawing from a Minimal Data Set Using a Single Wearable Sensor

The feasibility of prediction of same-limb kinematics using a single inertial measurement unit attached to the same limb has been demonstrated using machine learning. This study was performed to see if a single inertial measurement unit attached to the tibia can predict the opposite leg’s kinematics (cross-leg prediction). It also investigated if there is a minimal or smaller data set in a convolutional neural network model to predict lower extremity running kinematics under other running conditions and with what accuracy for the intra- and inter-participant situations. Ten recreational runners completed running exercises under five conditions, including treadmill running at speeds of 2, 2.5, 3, and 3.5 m/s and level-ground running at their preferred speed. A one-predict-all scheme was adopted to determine which running condition could be used to best predict a participant’s overall running kinematics. Running kinematic predictions were performed for intra- and inter-participant scenarios. Among the tested running conditions, treadmill running at 3 m/s was found to be the optimal condition for accurately predicting running kinematics under other conditions, with R2 values ranging from 0.880 to 0.958 and 0.784 to 0.936 for intra- and inter-participant scenarios, respectively. The feasibility of cross-leg prediction was demonstrated but with significantly lower accuracy than the same leg. The treadmill running condition at 3 m/s showed the highest intra-participant cross-leg prediction accuracy. This study proposes a novel, deep-learning method for predicting running kinematics under different conditions on a small training data set.

Ergonomic Design and Performance Evaluation of H-Suit for Human Walking

A soft exoskeleton for the hip flexion, named H-Suit, is developed to improve the walking endurance of lower limbs, delay muscle fatigue and reduce the activation level of hip flexors. Based on the kinematics and biomechanics of the hip joints, the ergonomic design of the H-Suit system is clearly presented and the prototype was developed. The profile of the auxiliary forces is planned in the auxiliary range where the forces start at the minimum hip angle, reach the maximum (120 N) and end at 90% of each gait cycle. The desired displacements of the traction unit which consist of the natural and elastic displacements of the steel cables are obtained by the experimental method. An assistance strategy is proposed to track the profile of the auxiliary forces by dynamically adjusting the compensation displacement Lc and the hold time Δt. The influences of the variables Lc and Δt on the natural gaits and auxiliary forces have been revealed and analyzed. The real profile of the auxiliary forces can be obtained and is consistent with the theoretical one by the proposed assistance strategy. The H-Suit without the drive unit has little effect on the EMG signal of the lower limbs. In the powered condition, the H-Suit can delay the muscle fatigue of the lower limbs. The average rectified value (ARV) slope decreases and the median frequency (MNF) slope increases significantly. Wearing the H-Suit resulted in a significant reduction of the vastus lateralis effort, averaged over subjects and walking speeds, of 13.3 ± 2.1% (p = 2 × 10−5).

Predicting the Internal Knee Abduction Impulse During Walking Using Deep Learning

Knee joint moments are commonly calculated to provide an indirect measure of knee joint loads. A shortcoming of inverse dynamics approaches is that the process of collecting and processing human motion data can be time-consuming. This study aimed to benchmark five different deep learning methods in using walking segment kinematics for predicting internal knee abduction impulse during walking. Three-dimensional kinematic and kinetic data used for the present analyses came from a publicly available dataset on walking (participants n = 33). The outcome for prediction was the internal knee abduction impulse over the stance phase. Three-dimensional (3D) angular and linear displacement, velocity, and acceleration of the seven lower body segment’s center of mass (COM), relative to a fixed global coordinate system were derived and formed the predictor space (126 time-series predictors). The total number of observations in the dataset was 6,737. The datasets were split into training (75%, n = 5,052) and testing (25%, n = 1685) datasets. Five deep learning models were benchmarked against inverse dynamics in quantifying knee abduction impulse. A baseline 2D convolutional network model achieved a mean absolute percentage error (MAPE) of 10.80%. Transfer learning with InceptionTime was the best performing model, achieving the best MAPE of 8.28%. Encoding the time-series as images then using a 2D convolutional model performed worse than the baseline model with a MAPE of 16.17%. Time-series based deep learning models were superior to an image-based method when predicting knee abduction moment impulse during walking. Future studies looking to develop wearable technologies will benefit from knowing the optimal network architecture, and the benefit of transfer learning for predicting joint moments.

DeepBBWAE-Net: A CNN-RNN Based Deep SuperLearner For Estimating Lower Extremity Sagittal Plane Joint Kinematics Using Shoe-Mounted IMU Sensors In Daily Living

Measurement of human body movement is an essential step in biomechanical analysis. The current standard for human motion capture systems uses infrared cameras to track reflective markers placed on the subject. While these systems can accurately track joint kinematics, the analyses are spatially limited to the lab environment. Though Inertial Measurement Unit (IMU) can eliminate the spatial limitations of the motion capture system, those systems are impractical for use in daily living due to the need for many sensors, typically one per body segment. Due to the need for practical and accurate estimation of joint kinematics, this study implements a reduced number of IMU sensors and employs machine learning algorithm to map sensor data to joint angles. Our developed algorithm estimates hip, knee, and ankle angles in the sagittal plane using two shoe-mounted IMU sensors in different practical walking conditions: treadmill, level overground, stair, and slope conditions. Specifically, we proposed five deep learning networks that use combinations of Convolutional Neural Networks (CNN) and Gated Recurrent Unit (GRU) based Recurrent Neural Networks (RNN) as base learners for our framework. Using those five baseline models, we proposed a novel framework, DeepBBWAE-Net, that implements ensemble techniques such as bagging, boosting, and weighted averaging to improve kinematic predictions. DeepBBWAE-Net predicts joint kinematics for the three joint angles under all the walking conditions with a Root Mean Square Error (RMSE) 6.93-29.0% lower than base models individually. This is the first study that uses a reduced number of IMU sensors to estimate kinematics in multiple walking environments.

Inertial Motion Capture-Based Wearable Systems for Estimation of Joint Kinetics: A Systematic Review

In biomechanics, joint kinetics has an important role in evaluating the mechanical load of the joint and understanding its motor function. Although an optical motion capture (OMC) system has mainly been used to evaluate joint kinetics in combination with force plates, inertial motion capture (IMC) systems have recently been emerging in joint kinetic analysis due to their wearability and ubiquitous measurement capability. In this regard, numerous studies have been conducted to estimate joint kinetics using IMC-based wearable systems. However, these have not been comprehensively addressed yet. Thus, the aim of this review is to explore the methodology of the current studies on estimating joint kinetic variables by means of an IMC system. From a systematic search of the literature, 48 studies were selected. This paper summarizes the content of the selected literature in terms of the (i) study characteristics, (ii) methodologies, and (iii) study results. The estimation methods of the selected studies are categorized into two types: the inverse dynamics-based method and the machine learning-based method. While these two methods presented different characteristics in estimating the kinetic variables, it was demonstrated in the literature that both methods could be applied with good performance for the kinetic analysis of joints in different daily activities.

Effects of Individualized Gait Rehabilitation Robotics for Gait Training on Hemiplegic Patients: Before-After Study in the Same Person

Background

Lower-limb exoskeleton robots are being widely used in gait rehabilitation training for patients with stroke. However, most of the current rehabilitation robots are guided by predestined gait trajectories, which are often different from the actual gait trajectories of specific patients. One solution is to train patients using individualized gait trajectories generated from the physical parameters of patients. Hence, we aimed to explore the effect of individual gaits on energy consumption situations during gait rehabilitation training for hemiplegic patients with lower-limb exoskeleton robots.

Methods

A total of 9 unilateral-hemiplegic patients were recruited for a 2-day experiment. On the first day of the experiment, the 9 patients were guided by a lower-limb exoskeleton robot, walking on flat ground for 15 min in general gait trajectory, which was gained by clinical gait analysis (CGA) method. On the other day, the same 9 patients wore the identical robot and walked on the same flat ground for 15 min in an individualized gait trajectory. The main physiological parameters including heart rate (HR) and peripheral capillary oxygen saturation (SpO2) were acquired via cardio tachometer and oximeter before and after the walking training. The energy consumption situation was indicated by the variation of the value of HR and SpO2 after walking training compared to before.

Results

Between-group comparison showed that the individualized gait trajectory training resulted in an increase in HR levels and a decrease in SpO2 levels compared to the general gait trajectory training. The resulting difference had a statistical significance of p < 0.05.

Conclusion

Using individualized gait guidance in rehabilitation walking training can significantly improve energy efficiency for hemiplegic patients with stroke.

Existing predictive methods applied to gait analysis of patients with diabetes: study protocol for a systematic review

INTRODUCTION

Type 2 diabetes can lead to gait abnormalities, including a longer stance phase, shorter steps and improper foot pressure distribution. Quantitative data from objective methods for evaluating gait patterns are accurate and cost-effective. In addition, it can also help predictive methods to forecast complications and develop early strategies to guide treatments. To date, no research has systematically summarised the predictive methods used to assess type 2 diabetic gait. Therefore, this protocol aims to identify which predictive methods have been employed to assess the diabetic gait.

METHODS AND ANALYSIS

This protocol will follow the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocol (PRISMA-P) statement. Electronic searches of articles from inception to January 2022 will be performed, from May 2021 to 31 January 2022, in the Web of Science, MEDLINE, Embase, IEEE Xplore Digital Library, Scopus, CINAHL, Google Scholar, APA PsycInfo, the Cochrane Library and in references of key articles and grey literature without language restrictions. We will include studies that examined the development and/or validation of predictive methods to assess type 2 diabetic gait in adults aged >18 years without amputations, use of assistive devices, ulcers or neuropathic pain. Two independent reviewers will screen the included studies and extract the data using a customised charting form. A third reviewer will resolve any disagreements. A narrative synthesis will be performed for the included studies. Risk of bias and quality of evidence will be assessed using the Prediction Model Risk of Bias Assessment Tool and the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis.

ETHICS AND DISSEMINATION

Ethical approval is not required because only available secondary published data will be analysed. The findings will be disseminated through peer-reviewed journals and/or presentations at relevant conferences and other media platforms.

PROSPERO REGISTRATION NUMBER

CDR42020199495.

Predicting Knee Joint Kinematics from Wearable Sensor Data in People with Knee Osteoarthritis and Clinical Considerations for Future Machine Learning Models

Deep learning models developed to predict knee joint kinematics are usually trained on inertial measurement unit (IMU) data from healthy people and only for the activity of walking. Yet, people with knee osteoarthritis have difficulties with other activities and there are a lack of studies using IMU training data from this population. Our objective was to conduct a proof-of-concept study to determine the feasibility of using IMU training data from people with knee osteoarthritis performing multiple clinically important activities to predict knee joint sagittal plane kinematics using a deep learning approach. We trained a bidirectional long short-term memory model on IMU data from 17 participants with knee osteoarthritis to estimate knee joint flexion kinematics for phases of walking, transitioning to and from a chair, and negotiating stairs. We tested two models, a double-leg model (four IMUs) and a single-leg model (two IMUs). The single-leg model demonstrated less prediction error compared to the double-leg model. Across the different activity phases, RMSE (SD) ranged from 7.04° (2.6) to 11.78° (6.04), MAE (SD) from 5.99° (2.34) to 10.37° (5.44), and Pearson’s R from 0.85 to 0.99 using leave-one-subject-out cross-validation. This study demonstrates the feasibility of using IMU training data from people who have knee osteoarthritis for the prediction of kinematics for multiple clinically relevant activities.

How Artificial Intelligence and Machine Learning Is Assisting Us to Extract Meaning from Data on Bone Mechanics?

Dramatic advancements in interdisciplinary research with the fourth paradigm of science, especially the implementation of computer science, nourish the potential for artificial intelligence (AI), machine learning (ML), and artificial neural network (ANN) algorithms to be applied to studies concerning mechanics of bones. Despite recent enormous advancement in techniques, gaining deep knowledge to find correlations between bone shape, material, mechanical, and physical responses as well as properties is a daunting task. This is due to both complexity of the material itself and the convoluted shapes that this complex material forms. Moreover, many uncertainties and ambiguities exist concerning the use of traditional computational techniques that hinders gaining a full comprehension of this advanced biological material. This book chapter offers a review of literature on the use of AI, ML, and ANN in the study of bone mechanics research. A main question as to why to implement AI and ML in the mechanics of bones is fully addressed and explained. This chapter also introduces AI and ML and elaborates on the main features of ML algorithms such as learning paradigms, subtypes, main ideas with examples, performance metrics, training algorithms, and training datasets. As a frequently employed ML algorithm in bone mechanics, feedforward ANNs are discussed to make their taxonomy and working principles more readily comprehensible to researchers. A summary as well as detailed review of papers that employed ANNs to learn from collected data on bone mechanics are presented. Reviewing literature on the use of these data-driven tools is essential since their wider application has the potential to: improve clinical assessments enabling real-time simulations; avoid and/or minimize injuries; and, encourage early detection of such injuries in the first place.

Prediction of Lower Extremity Multi-Joint Angles during Overground Walking by Using a Single IMU with a Low Frequency Based on an LSTM Recurrent Neural Network

The joint angle during gait is an important indicator, such as injury risk index, rehabilitation status evaluation, etc. To analyze gait, inertial measurement unit (IMU) sensors have been used in studies and continuously developed; however, they are difficult to utilize in daily life because of the inconvenience of having to attach multiple sensors together and the difficulty of long-term use due to the battery consumption required for high data sampling rates. To overcome these problems, this study propose a multi-joint angle estimation method based on a long short-term memory (LSTM) recurrent neural network with a single low-frequency (23 Hz) IMU sensor. IMU sensor data attached to the lateral shank were measured during overground walking at a self-selected speed for 30 healthy young persons. The results show a comparatively good accuracy level, similar to previous studies using high-frequency IMU sensors. Compared to the reference results obtained from the motion capture system, the estimated angle coefficient of determination (R2) is greater than 0.74, and the root mean square error and normalized root mean square error (NRMSE) are less than 7° and 9.87%, respectively. The knee joint showed the best estimation performance in terms of the NRMSE and R2 among the hip, knee, and ankle joints.

Machine Learning Model to Estimate Net Joint Moments during Lifting Task Using Wearable Sensors: A Preliminary Study for Design of Exoskeleton Control System

Accurately measuring the lower extremities and L5/S1 moments is important since L5/S1 moments are the principal parameters that measure the risk of musculoskeletal diseases during lifting. In this study, protocol that predicts lower extremities and L5/S1 moments with an insole sensor was proposed to replace the prior methods that have spatial constraints. The protocol is hierarchically composed of a classification model and a regression model to predict joint moments. Additionally, a single LSTM model was developed to compare with proposed protocol. To optimize hyperparameters of the machine learning model and input feature, Bayesian optimization method was adopted. As a result, the proposed protocol showed a relative root mean square error (rRMSE) of 8.06~13.88% while the single LSTM showed 9.30~18.66% rRMSE. This protocol in this research is expected to be a starting point for developing a system for estimating the lower extremity and L5/S1 moment during lifting that can replace the complex prior method and adopted to workplace environments. This novel study has the potential to precisely design a feedback iterative control system of an exoskeleton for the appropriate generation of an actuator torque.

A Deep Learning Approach for Foot Trajectory Estimation in Gait Analysis Using Inertial Sensors

Gait performance is an important marker of motor and cognitive decline in older adults. An instrumented gait analysis resorting to inertial sensors allows the complete evaluation of spatiotemporal gait parameters, offering an alternative to laboratory-based assessments. To estimate gait parameters, foot trajectories are typically obtained by integrating acceleration two times. However, to deal with cumulative integration errors, additional error handling strategies are required. In this study, we propose an alternative approach based on a deep recurrent neural network to estimate heel and toe trajectories. We propose a coordinate frame transformation for stride trajectories that eliminates the dependency from previous strides and external inputs. Predicted trajectories are used to estimate an extensive set of spatiotemporal gait parameters. We evaluate the results in a dataset comprising foot-worn inertial sensor data acquired from a group of young adults, using an optical motion capture system as a reference. Heel and toe trajectories are predicted with low errors, in line with reference trajectories. A good agreement is also achieved between the reference and estimated gait parameters, in particular when turning strides are excluded from the analysis. The performance of the method is shown to be robust to imperfect sensor-foot alignment conditions.

The Use of Synthetic IMU Signals in the Training of Deep Learning Models Significantly Improves the Accuracy of Joint Kinematic Predictions

Gait analysis based on inertial sensors has become an effective method of quantifying movement mechanics, such as joint kinematics and kinetics. Machine learning techniques are used to reliably predict joint mechanics directly from streams of IMU signals for various activities. These data-driven models require comprehensive and representative training datasets to be generalizable across the movement variability seen in the population at large. Bottlenecks in model development frequently occur due to the lack of sufficient training data and the significant time and resources necessary to acquire these datasets. Reliable methods to generate synthetic biomechanical training data could streamline model development and potentially improve model performance. In this study, we developed a methodology to generate synthetic kinematics and the associated predicted IMU signals using open source musculoskeletal modeling software. These synthetic data were used to train neural networks to predict three degree-of-freedom joint rotations at the hip and knee during gait either in lieu of or along with previously measured experimental gait data. The accuracy of the models’ kinematic predictions was assessed using experimentally measured IMU signals and gait kinematics. Models trained using the synthetic data out-performed models using only the experimental data in five of the six rotational degrees of freedom at the hip and knee. On average, root mean square errors in joint angle predictions were improved by 38% at the hip (synthetic data RMSE: 2.3°, measured data RMSE: 4.5°) and 11% at the knee (synthetic data RMSE: 2.9°, measured data RMSE: 3.3°), when models trained solely on synthetic data were compared to measured data. When models were trained on both measured and synthetic data, root mean square errors were reduced by 54% at the hip (measured + synthetic data RMSE: 1.9°) and 45% at the knee (measured + synthetic data RMSE: 1.7°), compared to measured data alone. These findings enable future model development for different activities of clinical significance without the burden of generating large quantities of gait lab data for model training, streamlining model development, and ultimately improving model performance.

Kinematics, Speed, and Anthropometry-Based Ankle Joint Torque Estimation: A Deep Learning Regression Approach

Powered Assistive Devices (PADs) have been proposed to enable repetitive, user-oriented gait rehabilitation. They may include torque controllers that typically require reference joint torque trajectories to determine the most suitable level of assistance. However, a robust approach able to automatically estimate user-oriented reference joint torque trajectories, namely ankle torque, while considering the effects of varying walking speed, body mass, and height on the gait dynamics, is needed. This study evaluates the accuracy and generalization ability of two Deep Learning (DL) regressors (Long-Short Term Memory and Convolutional Neural Network (CNN)) to generate user-oriented reference ankle torque trajectories by innovatively customizing them according to the walking speed (ranging from 1.0 to 4.0 km/h) and users’ body height and mass (ranging from 1.51 to 1.83 m and 52.0 to 83.7 kg, respectively). Furthermore, this study hypothesizes that DL regressors can estimate joint torque without resourcing electromyography signals. CNN was the most robust algorithm (Normalized Root Mean Square Error: 0.70 ± 0.06; Spearman Correlation: 0.89 ± 0.03; Coefficient of Determination: 0.91 ± 0.03). No statistically significant differences were found in CNN accuracy (p-value > 0.05) whether electromyography signals are included as inputs or not, enabling a less obtrusive and accurate setup for torque estimation.

A Comparison of Three Neural Network Approaches for Estimating Joint Angles and Moments from Inertial Measurement Units

The application of artificial intelligence techniques to wearable sensor data may facilitate accurate analysis outside of controlled laboratory settings—the holy grail for gait clinicians and sports scientists looking to bridge the lab to field divide. Using these techniques, parameters that are difficult to directly measure in-the-wild, may be predicted using surrogate lower resolution inputs. One example is the prediction of joint kinematics and kinetics based on inputs from inertial measurement unit (IMU) sensors. Despite increased research, there is a paucity of information examining the most suitable artificial neural network (ANN) for predicting gait kinematics and kinetics from IMUs. This paper compares the performance of three commonly employed ANNs used to predict gait kinematics and kinetics: multilayer perceptron (MLP); long short-term memory (LSTM); and convolutional neural networks (CNN). Overall high correlations between ground truth and predicted kinematic and kinetic data were found across all investigated ANNs. However, the optimal ANN should be based on the prediction task and the intended use-case application. For the prediction of joint angles, CNNs appear favourable, however these ANNs do not show an advantage over an MLP network for the prediction of joint moments. If real-time joint angle and joint moment prediction is desirable an LSTM network should be utilised.

Using deep neural networks for kinematic analysis: Challenges and opportunities

Kinematic analysis is often performed in a lab using optical cameras combined with reflective markers. With the advent of artificial intelligence techniques such as deep neural networks, it is now possible to perform such analyses without markers, making outdoor applications feasible. In this paper I summarise 2D markerless approaches for estimating joint angles, highlighting their strengths and limitations. In computer science, so-called "pose estimation" algorithms have existed for many years. These methods involve training a neural network to detect features (e.g. anatomical landmarks) using a process called supervised learning, which requires "training" images to be manually annotated. Manual labelling has several limitations, including labeller subjectivity, the requirement for anatomical knowledge, and issues related to training data quality and quantity. Neural networks typically require thousands of training examples before they can make accurate predictions, so training datasets are usually labelled by multiple people, each of whom has their own biases, which ultimately affects neural network performance. A recent approach, called transfer learning, involves modifying a model trained to perform a certain task so that it retains some learned features and is then re-trained to perform a new task. This can drastically reduce the required number of training images. Although development is ongoing, existing markerless systems may already be accurate enough for some applications, e.g. coaching or rehabilitation. Accuracy may be further improved by leveraging novel approaches and incorporating realistic physiological constraints, ultimately resulting in low-cost markerless systems that could be deployed both in and outside of the lab.

Machine learning for rapid estimation of lower extremity muscle and joint loading during activities of daily living

Joint contact and muscle forces estimated with musculoskeletal modeling techniques offer useful metrics describing movement quality that benefit multiple research and clinical applications. The expensive processing of laboratory data associated with generating these outputs presents challenges to researchers and clinicians, including significant time and expertise requirements that limit the number of subjects typically evaluated. The objective of the current study was to develop and compare machine learning techniques for rapid, data-driven estimation of musculoskeletal metrics from derived gait lab data. OpenSim estimates of patient joint and muscle forces during activities of daily living were simulated using laboratory data from 70 total knee replacement patients and used to develop 4 different machine learning algorithms. Trained machine learning models predicted both trend and magnitude of estimated joint contact (mean correlation coefficients ranging from 0.93 to 0.94 during gait) and muscle forces (mean correlation coefficients ranging from 0.83 to 0.91 during gait) based on anthropometrics, ground reaction forces, and joint angle data. Patient mechanics were accurately predicted by recurrent neural networks, even after removing dependence on key subsets of predictor features. The ability to quickly estimate patient mechanics from derived measurements of movement has the potential to broaden the impact of musculoskeletal modeling by enabling faster assessment in both clinical and research settings.

Length of hospital stay prediction with an integrated approach of statistical-based fuzzy cognitive maps and artificial neural networks

Chronic obstructive pulmonary disease (COPD) is a global burden, which is estimated to be the third leading cause of death worldwide by 2030. The economic burden of COPD grows continuously because it is not a curable disease. These conditions make COPD an important research field of artificial intelligence (AI) techniques in medicine. In this study, an integrated approach of the statistical-based fuzzy cognitive maps (SBFCM) and artificial neural networks (ANN) is proposed for predicting length of hospital stay of patients with COPD, who admitted to the hospital with an acute exacerbation. The SBFCM method is developed to determine the input variables of the ANN model. The SBFCM conducts statistical analysis to prepare preliminary information for the experts and then collects expert opinions accordingly, to define a conceptual map of the system. The integration of SBFCM and ANN methods provides both statistical data and expert opinion in the prediction model. In the numerical application, the proposed approach outperformed the conventional approach and other machine learning algorithms with 79.95% accuracy, revealing the power of expert opinion involvement in medical decisions. A medical decision support framework is constructed for better prediction of length of hospital stay and more effective hospital management.

Real-Time Prediction of Joint Forces by Motion Capture and Machine Learning

Conventional biomechanical modelling approaches involve the solution of large systems of equations that encode the complex mathematical representation of human motion and skeletal structure. To improve stability and computational speed, being a common bottleneck in current approaches, we apply machine learning to train surrogate models and to predict in near real-time, previously calculated medial and lateral knee contact forces (KCFs) of 54 young and elderly participants during treadmill walking in a speed range of 3 to 7 km/h. Predictions are obtained by fusing optical motion capture and musculoskeletal modeling-derived kinematic and force variables, into regression models using artificial neural networks (ANNs) and support vector regression (SVR). Training schemes included either data from all subjects (LeaveTrialsOut) or only from a portion of them (LeaveSubjectsOut), in combination with inclusion of ground reaction forces (GRFs) in the dataset or not. Results identify ANNs as the best-performing predictor of KCFs, both in terms of Pearson R (0.89-0.98 for LeaveTrialsOut and 0.45-0.85 for LeaveSubjectsOut) and percentage normalized root mean square error (0.67-2.35 for LeaveTrialsOut and 1.6-5.39 for LeaveSubjectsOut). When GRFs were omitted from the dataset, no substantial decrease in prediction power of both models was observed. Our findings showcase the strength of ANNs to predict simultaneously multi-component KCF during walking at different speeds-even in the absence of GRFs-particularly applicable in real-time applications that make use of knee loading conditions to guide and treat patients.

Prediction of ground reaction force and joint moments based on optical motion capture data during gait

The standard camera- and force plate-based set-up for motion analysis suffers from the disadvantage of being limited to laboratory settings. Since adaptive algorithms are able to learn the connection between known inputs and outputs and generalise this knowledge to unknown data, these algorithms can be used to leverage motion analysis outside the laboratory. In most biomechanical applications, feedforward neural networks are used, although these networks can only work on time normalised data, while recurrent neural networks can be used for real time applications. Therefore, this study compares the performance of these two kinds of neural networks on the prediction of ground reaction force and joint moments of the lower limbs during gait based on joint angles determined by optical motion capture as input data. The accuracy of both networks when generalising to new data was assessed using the normalised root-mean-squared error, the root-mean-squared error and the correlation coefficient as evaluation metrics. Both neural networks demonstrated a high performance and good capabilities to generalise to new data. The mean prediction accuracy over all parameters applying a feedforward network was higher (r = 0.963) than using a recurrent long short-term memory network (r = 0.935).

Artificial Neural Networks in Motion Analysis-Applications of Unsupervised and Heuristic Feature Selection Techniques

The use of machine learning to estimate joint angles from inertial sensors is a promising approach to in-field motion analysis. In this context, the simplification of the measurements by using a small number of sensors is of great interest. Neural networks have the opportunity to estimate joint angles from a sparse dataset, which enables the reduction of sensors necessary for the determination of all three-dimensional lower limb joint angles. Additionally, the dimensions of the problem can be simplified using principal component analysis. Training a long short-term memory neural network on the prediction of 3D lower limb joint angles based on inertial data showed that three sensors placed on the pelvis and both shanks are sufficient. The application of principal component analysis to the data of five sensors did not reveal improved results. The use of longer motion sequences compared to time-normalised gait cycles seems to be advantageous for the prediction accuracy, which bridges the gap to real-time applications of long short-term memory neural networks in the future.

CNN-Based Estimation of Sagittal Plane Walking and Running Biomechanics From Measured and Simulated Inertial Sensor Data

Machine learning is a promising approach to evaluate human movement based on wearable sensor data. A representative dataset for training data-driven models is crucial to ensure that the model generalizes well to unseen data. However, the acquisition of sufficient data is time-consuming and often infeasible. We present a method to create realistic inertial sensor data with corresponding biomechanical variables by 2D walking and running simulations. We augmented a measured inertial sensor dataset with simulated data for the training of convolutional neural networks to estimate sagittal plane joint angles, joint moments, and ground reaction forces (GRFs) of walking and running. When adding simulated data, the root mean square error (RMSE) of the test set of hip, knee, and ankle joint angles decreased up to 17%, 27% and 23%, the RMSE of knee and ankle joint moments up to 6% and the RMSE of anterior-posterior and vertical GRF up to 2 and 6%. Simulation-aided estimation of joint moments and GRFs was limited by inaccuracies of the biomechanical model. Improving the physics-based model and domain adaptation learning may further increase the benefit of simulated data. Future work can exploit biomechanical simulations to connect different data sources in order to create representative datasets of human movement. In conclusion, machine learning can benefit from available domain knowledge on biomechanical simulations to supplement cumbersome data collections.